PENGENALAN POLA GARIS DASAR KALIMAT PADA TULISAN TANGAN UNTUK MENGETAHUI KARAKTER SESEORANG DENGAN MENGGUNAKAN ALGORITMA RESILIENT BACKPROPAGATION ABSTRAK Juventus Suharta (0722026) Jurusan Teknik Elektro Universitas Kristen Maranatha email :
[email protected] Grafologi adalah ilmu yang mempelajari karakter seseorang seseorang dengan cara menganalisa tulisan tangan. Menganalisa tulisan tangan sangatlah membantu dalam banyak bidang saat ini, misalnya dalam bidang pendidikan, kriminalitas dan dapat digunakan sebagai konseling. Salah satu cara yang digunakan dalam grafologi untuk mengetahui karakter seseorang, adalah dengan menganalisa pola garis dasar kalimat dari tulisan tangan. Pada Tugas Akhir ini
dirancang dan direalisasikan perangkat lunak
berbasis Jaringan Saraf Tiruan untuk mengenali pola garis dasar kalimat dari tulisan tangan manusia, dengan menggunakan nilai rata-rata dari posisi pixel yang bernilai 1 pada citra yang akan menjadi masukan dari data latih dan data uji pada Algoritma Resilient Backpropagation. Perangkat lunak ini direalisasikan menggunakan MATLAB R2008a. Perangkat lunak pengenalan pola garis dasar tulisan tangan pada Tugas Akhir ini berhasil direalisasikan dan diperoleh keberhasilan pengenalan sebesar 65,63% pada pengujian. Persentase pengenalan masing-masing pola: untuk pola garis lurus 75%, pola garis menaik 62,5%, untuk pola garis menurun, dan untuk pola garis acak 87,5%.
Kata Kunci : Grafologi, Jaringan Saraf Tiruan, Resilient Backpropagation, Pengenalan Pola Garis Dasar Tulisan Tangan. i Universitas Kristen Maranatha
HANDWRITING BASELINE PATTERN RECOGNITION TO IDENTIFY HUMAN CHARACTER USING RESILIENT BACKPROPAGATION ALGORITHM
ABSTRACT Juventus Suharta (0722026) Department of Electrical Engineering Maranatha Christian University email :
[email protected]
Graphology is the study of a person's character by analyzing handwriting. Analyzing handwriting is helpful in many areas today, for example in education, crime and can be used as counseling. One method used in graphology to know the character of a person, is by analyzing the baseline pattern of handwritten sentences.
This final project is designed and realized a software based Artificial
Neural Networks to recognize patterns of baseline sentences of human handwriting, by using out the average value of the positions of pixels of value 1 in entire image that will become input from the training data and testing data in Resilient Backpropagation Algorithm. The software is realized using MATLAB R2008a. Handwriting Baseline Pattern Recognition on this final project successfully realized. In testing, this software has 65.63% recognize well. Percentage identity of each pattern: for straight line pattern 75%, for ascending line pattern 62,5%, for descending line pattern, and for random line pattern 87,5%.
Keywords :
Graphology, Neural Network, Resilient
Backpropagation,
Handwriting Baseline Pattern Recognition.
ii Universitas Kristen Maranatha
DAFTAR ISI ABSTRAK ............................................................................................................... i ABSTRACT ............................................................................................................ ii KATA PENGANTAR ............................................................................................ ii DAFTAR GAMBAR ........................................................................................... viii DAFTAR TABEL ....................................................................................................x BAB I .......................................................................................................................1 Pendahuluan .............................................................................................................1 I.1 Latar Belakang ................................................................................................1 I.2 Identifikasi Masalah........................................................................................2 I.3 Pembatasan Masalah .......................................................................................2 I.4 Tujuan Tugas Akhir ........................................................................................3 I.5 Sistematika Penelitian .....................................................................................3 BAB II ......................................................................................................................5 LANDASAN TEORI ...............................................................................................5 II.1 Teori Dasar Citra Dijital ................................................................................5 II.1.1 Pixel Dan Resolusi Citra ........................................................................5 II.1.2 Ciri Citra .................................................................................................6 II.2 Pengolahan Citra Dijital ................................................................................6 II.2.1 Grayscale................................................................................................6 II.2.2 Binerisasi ................................................................................................7 II.2.3 Cropping .................................................................................................8 II.2.4 Resizing...................................................................................................8 II.3 Jaringan Syaraf Biologi .................................................................................8 II.4 Jaringan Syaraf Tiruan ................................................................................10 II.4.1 Arsitektur Jaringan Syaraf Tiruan ........................................................10 II.4.2 Fungsi Aktivasi .....................................................................................12 II.4.3 Bias .......................................................................................................13 II.4.4 Laju Pembelajaran / Learning Rate ......................................................13 II.5 Metode Backpropagation ............................................................................14 II.5.1 Arsitektur Backpropagation .................................................................15 II.5.2 Algoritma Pelatihan Backpropagation .................................................17 II.5.3 Inisialisasi Bobot Awal Dan Bias .........................................................19 iii Universitas Kristen Maranatha
II.5.3.1 Inisialisasi Acak.............................................................................20 II.5.3.2 Inisialisasi Nguyen-Widrow ..........................................................20 II.5.4 Perhitungan Error.................................................................................21 II.5.4 Pengujian ..............................................................................................22 II.6 Metode Resilient Backpropagation (Rprop) ...............................................23 II.6.1 Arsitektur Resilient Backpropagation (Rprop) .....................................24 II.6.2 Algoritma Resilient Backpropagation (Rprop) ...................................24 II.7 Grafologi......................................................................................................33 II.7.1 Definisi Grafologi .................................................................................33 II.7.2 Pola Garis Dasar kalimat Dalam Grafologi .........................................34 BAB III...................................................................................................................37 PERANCANGAN PERANGKAT LUNAK .........................................................37 III.1 Arsitektur Perancangan ..............................................................................37 III.2 Diagram Alir ..............................................................................................38 III.2.1 Diagram Alir Praproses .......................................................................39 III.2.2 Diagram Alir Tahapan Mencari Nilai Rata-rata..................................40 III.2.3 Diagram Alir Pelatihan Algoritma Resilient Backpropagation ..........43 III.2.4 Diagram Alir Pengujian Algoritma Resilient Backpropagation..........44 III.3 Penerapan Grafologi...................................................................................45 III.4 Perancangan Antarmuka Pemakai (User Interface)...................................48 BAB IV ..................................................................................................................50 SIMULASI DAN ANALISA.................................................................................50 IV.1 Proses Pelatihan .........................................................................................50 IV.1.1 Percobaan 1: Pengaruh Jumlah Hidden Neuron .................................50 IV.1.2 Percobaan 2: Pengaruh Faktor Naik (FN) ...........................................52 IV.1.3 Percobaan 3: Pengaruh Faktor Turun (FT) .........................................54 IV.2 Proses Pengujian ........................................................................................55 IV.2.1 Pengujian I ..........................................................................................59 IV.2.2 Pengujian II .........................................................................................63 IV.3 Analisa .......................................................................................................67 BAB V....................................................................................................................68 KESIMPULAN DAN SARAN ..............................................................................68 V.1 Kesimpulan..................................................................................................68 V.2 Saran ............................................................................................................68 iv Universitas Kristen Maranatha
DAFTAR PUSTAKA ............................................................................................69 LAMPIRAN A ........................................................................................................ A LAMPIRAN B ........................................................................................................ B LAMPIRAN C ........................................................................................................ C LAMPIRAN D ........................................................................................................ D
v Universitas Kristen Maranatha
DAFTAR GAMBAR
Gambar 2. 1 Sel Syaraf Biologi ..............................................................................9 Gambar 2. 2 Single Layer Network .......................................................................11 Gambar 2. 3 Multi Layer Network ........................................................................11 Gambar 2. 4 Model JST Resilient Backpropagation ............................................24 Gambar 2. 5 Arsitektur Risilent Backpropagation Dalam Contoh Perhitungan ...29 Gambar 2. 6 Pola Garis Dasar Lurus ....................................................................34 Gambar 2. 7 Pola Garis Dasar Naik ......................................................................35 Gambar 2. 8 Pola Garis Dasar Turun ....................................................................35 Gambar 2. 9 Pola Garis Dasar Tidak Beraturan....................................................36 Gambar 3. 1
Arsitektur Resilient Backpropagation Pengenalan Pola Garis........... Dasar Tulisan ..................................................................................37
Gambar 3. 2
Diagram Alir Utama .......................................................................38
Gambar 3. 3
Diagram Alir Praproses ..................................................................39
Gambar 3. 4
Diagram Alir Tahapan Mencari Nilai Rata-rata.............................40
Gambar 3. 5
Citra 4x4 .........................................................................................41
Gambar 3. 6
Citra 4x4 Dengan Garis Nilai Rata-rata Posisi Pixel bernilai 1 .....42
Gambar 3. 7
Diagram Alir Pelatihan Algoritma Risilent Backpropagation .......43
Gambar 3. 8
Diagram Alir Pengujian Algoritma Risilent Backpropagation ......44
Gambar 3. 9
Pola Garis Dasar Lurus ..................................................................45
Gambar 3. 10 Pola Garis Dasar Naik ....................................................................46 Gambar 3. 11 Pola Garis Dasar Turun ..................................................................46 Gambar 3. 12 Pola Garis Dasar Tidak Beraturan..................................................47 Gambar 3. 13 Antarmuka Perangkat Lunak ..........................................................48
Gambar 4. 1 Grafik Perbandingan Jumlah Pengenalan Terhadap Jumlah................ Hidden Neuron ..................................................................................50 Gambar 4. 2 Grafik Perbandingan Jumlah Pengenalan Pola Terhadap FN ..........52 Gambar 4. 3 Grafik Perbandingan Jumlah Pengenalan Pola Terhadap FT...........54 vi Universitas Kristen Maranatha
Gambar 4. 4 Tampilan Hasil Pengujian Pola Garis Dasar Tulisan Tangan ..........55 Gambar 4. 5 Pengujian Pola Garis Dasar Lurus ...................................................56 Gambar 4. 6 Pengujian Pola Garis Dasar Naik .....................................................56 Gambar 4. 7 Pengujian Pola Garis Dasar Turun ...................................................57 Gambar 4. 8 Pengujian Pola Garis Dasar Acak ....................................................57
vii Universitas Kristen Maranatha
DAFTAR TABEL
Tabel 2. 1 Bobot Dari Input Layer Ke Hidden Layer ...........................................28 Tabel 2. 2 Bobot Dari Hidden Layer Ke Output Layer.........................................28
Tabel 3. 1 Atribut MATLAB Pada Perancangan Perangkat Lunak ......................48
Tabel 4. 1 Tabel Percobaan Jumlah Hidden Neuron.............................................51 Tabel 4. 2 Tabel Percobaan Faktor Naik ...............................................................53 Tabel 4. 3 Tabel Percobaan Faktor Turun .............................................................54 Tabel 4. 4 Tabel Data Input Pengujian..................................................................58 Tabel 4. 5 Tabel Hasil Pengujian I ........................................................................59 Tabel 4. 6 Tabel Hasil Pengujian II ......................................................................63
viii Universitas Kristen Maranatha
